Records and AI data leakage: what a pasted care plan or case record gives away in ChatGPT or Claude.
Care plans, clinical records, treatment plans and sworn statements hold the most intimate detail a practice keeps about a named person: diagnoses, plans, decisions and history. Staff paste them into AI tools to summarise or rewrite them. These pages mark up real records across professions.
Accountants
Law firms
Financial advisers
Aged care
Diagnoses, medications, behaviours, mobility and the substitute decision-maker. The whole person, on four pages.
See what leaves → The behaviour support planThe triggers, the psychotropic, who consented, what was tried first. The record the law requires, about the most vulnerable resident in the home.
See what leaves → The palliative care planPrognosis, treatment limits, symptom plan, faith, a disagreement between her children and funeral wishes. The most private page in the file.
See what leaves →Mortgage brokers
Healthcare
General practitioners (GPs)
Conveyancers
Migration agents
NDIS providers
Diagnosis, medication, triggers, the lockbox and a relationship he has asked to keep private. The whole person, in one record.
See what leaves → The positive behaviour support planBehaviours, the reason for them, the locked cupboard, the PRN medication and who authorised it. The plan the Commission has warned providers about putting into AI tools.
See what leaves →Recruiters
Common questions
Is it safe to put a record into ChatGPT or Claude?
Do ChatGPT and Claude use the information in a record?
Can I remove the names from a record before using ChatGPT or Claude?
How can a practice use AI on records without sending them to ChatGPT or Claude?
What do the pages about records show?
Are the example records real?
Which professions are covered?
See Dominion answering from your own documents.
A demonstration takes 30 minutes, uses no client data, and comes with no obligation.
Book a demo